Visualizing Interacting Biomolecules In Situ

To my beloved family
”It is very easy to answer many of these fundamental biological questions;
you just look at the thing!”
Richard P. Feynman
List of Papers
This thesis is based on the following papers, which are referred to in the text
by their Roman numerals.
I
Jarvius M*, Paulsson J*, Weibrecht I*, Leuchowius KJ,
Andersson AC, Wählby C, Gullberg M, Botling J, Sjöblom T,
Markova B, Östman A, Landegren U, Söderberg O (2007) In
situ detection of phosphorylated platelet-derived growth factor
receptor beta using a generalized proximity ligation method.
Mol Cell Proteomics, 6(9): 1500-9
II
Leuchowius KJ, Weibrecht I, Landegren U, Gedda L,
Söderberg O (2009) Flow cytometric in situ proximity ligation
analyses of protein interactions and post-translational
modification of the epidermal growth factor receptor family.
Cytometry A 75(10): 833-9
III
Weibrecht I*, Grundberg I*, Nilsson M, Söderberg O
(manuscript) Simultaneous visualization of both signaling
cascade activity and end-point gene expression in single cells.
Accepted for publication in PLoS ONE
IV
Weibrecht I, Gavrilovic M, Lindbom L, Landegren U, Wählby
C, Söderberg O (manuscript) Visualizing individual sequencespecific protein-DNA interactions in situ. Submitted
*
equal contribution
Reprints were made with permission from the respective publishers.
Related Work by the Author
Peer Reviewed Research Articles
Mahmoudi S, Henriksson S, Weibrecht I, Smith S, Söderberg O, Strömblad
S, Wiman KG, Farnebo M (2010) WRAP53 is essential for Cajal body
formation and for targeting the survival of motor neuron complex to Cajal
bodies. PLoS Biol. 8(11): e1000521
Weibrecht I, Böhmer SA, Dagnell M, Kappert K, Östman A, Böhmer FD
(2007) Oxidation sensitivity of the catalytic cysteine of the protein-tyrosine
phosphatases SHP-1 and SHP-2. Free Radical Biol Med 43(1): 100-10
Review Articles
Leuchowius KJ, Weibrecht I, Söderberg O (2011) In situ proximity ligation
assay for microscopy and flow cytometry. Curr Protoc Cytom 9.36.1-16
Weibrecht I, Leuchowius KJ, Clausson CM, Conze T, Jarvius M, Howell
WM, Kamali-Moghaddam M, Söderberg O (2010) Proximity ligation
assays: a recent addition to the proteomics toolbox. Expert Rev Proteomics
7(3): 401-9
Söderberg O, Leuchowius KJ, Gullberg M, Jarvius M, Weibrecht I, Larsson
LG, Landegren U (2008) Characterizing proteins and their interactions in
cells and tissues using the in situ proximity ligation assay. Methods 45(3):
227-32
Book Chapter
Grundberg I*, Weibrecht I*, Landegren U (2010) Amplified single molecule
detection. Handbook of Nanophysics. Principles and methods, CRC Press,
Taylor & Francis Group, ISBN: 978-1-4200-7540-3, pp 40-1-11
*
equal contribution
Contents
Introduction – take a closer look ................................................................... 11
Methods to study protein-protein interactions in situ ............................... 13
Methods to study protein-DNA interactions in situ ................................. 20
Methods to study nucleic acids in situ...................................................... 24
Model systems .............................................................................................. 27
PDGFR and DUSP6 ............................................................................... 27
EGFR family ............................................................................................ 28
Chromatin structure and Alu-repeats ....................................................... 28
Present investigations.................................................................................... 31
Paper I: In situ detection of phosphorylated platelet-derived growth factor
receptor using a generalized proximity ligation method ....................... 31
Aim of the study .................................................................................. 31
Methods ............................................................................................... 31
Results ................................................................................................. 32
Discussion ............................................................................................ 33
Paper II: Flow cytometric in situ proximity ligation analyses of protein
interactions and post-translational modification of the epidermal growth
factor receptor family ............................................................................... 33
Aim of the study .................................................................................. 33
Methods ............................................................................................... 33
Results ................................................................................................. 34
Discussion ............................................................................................ 34
Paper III: Simultaneous visualization of both signaling cascade activity
and end-point gene expression in single cells. ......................................... 35
Aim of the study .................................................................................. 35
Methods ............................................................................................... 35
Results ................................................................................................. 36
Discussion ............................................................................................ 37
Paper IV: Visualizing individual, sequence-specific protein-DNA
interactions in situ .................................................................................... 37
Aim of the study .................................................................................. 37
Methods ............................................................................................... 38
Results ................................................................................................. 39
Discussion ............................................................................................ 40
Summary and future perspective................................................................... 41
Acknowledgments......................................................................................... 44
References ..................................................................................................... 49
Abbreviations
ACTB
BiFC
BJhTert
bp
BRET
cDNA
ChIP
DAPI
DNA
DUSP6
EGF(R)
EndoIV
ERK
FISH
FITC
FLIM
FRET
GFP
GRB2
H3K4 / 9
Hek293
HER2-4
IF
IgG
IHC
In situ PLA
ISH
Kb
LNA
m
MAPK
Mb
MCF-7
MEK
miRNA
MKP
-actin mRNA
Bimolecular fluorescence complementation
Immortalized human foreskin fibroblast cell line
Base pair
Bioluminescence resonance energy transfer
Complementary DNA
Chromatin immunoprecipitation
4',6-diamidino-2-phenylindole
Deoxyribonucleic acid
Dual specificity phosphatase 6 mRNA
Epidermal growth factor (receptor)
Endonuclease IV
Extracellular signal-regulated kinase
Fluorescence in situ hybridization
Fluorescein isothiocyanate
Fluorescence life time imaging
Förster (Fluorescence) resonance energy transfer
Green fluorescent protein
Growth factor receptor-bound protein 2
Histone H3 lysine 4 / 9
Human embryonic kidney 293 cell line
Human epidermal growth factor receptor 2-4
Immunofluorescence
Immunoglobulin G
Immunohistochemistry
In situ proximity ligation assay
In situ hybridization
Kilobase
Locked nucleic acid
Meter
Mitogen-activated protein kinase
Megabase
Human breast adenocarcinoma cell line
Mitogen-activated protein kinase kinase
MicroRNA
Mitogen-activated protein kinase phosphatase
mRNA
NIH3T3
nm
OLA
PAE /
PCA
PCR
PDGF-BB / -DD
PDGFR /
PDI
pEGFR
PI3K
PLA
PPI
PRINS
PTM
pY
qPCR
RCA
RCP
RNA
RTK
SDS-PAGE
SH2
SINE
SKOV-3
SNP
SOS
TNF
TSA
U343
UNG
VEGFR
Messenger RNA
Mouse embryonic fibroblast cell line
Nanometer
Oligonucleotide ligation assay
Porcine aortic endothelial cell line expressing PDGFR /
Protein fragment complementation assay
Polymerase chain reaction
Platelet-derived growth factor-BB / -DD
Platelet-derived growth factor receptor alpha / beta
Protein-DNA interaction
Phosphorylated EGFR
Phosphoinositide 3-kinase
Proximity ligation assay
Protein-protein interaction
Primed in situ labeling
Posttranslational modification
Phosphorylated tyrosine
Quantitative PCR
Rolling circle amplification
Rolling circle product
Ribonucleic acid
Receptor tyrosine kinase
Sodium dodecyl sulfate - polyacrylamide gel electrophoresis
Src homology 2
Short interspersed nuclear element
Human ovary adenocarcinoma cell line
Single nucleotide polymorphism
Son of Sevenless
Tumor necrosis factor alpha
Tyramide signal amplification
Human neuronal glioblastoma cell line
Uracil-DNA glycosylase
Vascular endothelial growth factor receptor
Introduction – take a closer look
The human body consists of about 1013 cells1, all containing the same ~3 x
109 bp of genomic DNA2, a ~2 m long molecule3 packed into the nucleus.
What information cells use from their genomic DNA, e.g. which genes and
miRNA are expressed, will differ substantially between cell types and the
expressed molecules span concentration ranges from only a couple of
molecules up to millions of them4. Complexity is further increased by taking
into account all intercellular communication that is required to sustain a
multicellular organism. All cells in a given population, even within clonal
expansions from each other – e.g. in immortalized cell cultures – are slightly
different because of position in cell cycle or access to stimuli from
surrounding cells. Even the molecules present in each cell vary: proteins
with different modifications, expressed mRNAs or accessibility of certain
DNA-sequences for transcription factors. When looking at this impressive
complexity, it seems amazing that an organism like the human body
functions properly most of the time, ensured by a concert of interacting
molecules, transporting signals within one cell and from one to another,
directing cells to e.g. division, differentiation or apoptosis. How cells cope
with this seemingly overwhelming amount of – sometimes contradictory –
information that they receive in every second of their lives can yet only be
partly appreciated. But not until we understand how processes in a cell and
body interweave and synergize, will we be able to efficiently treat diseases
that occur when this teamwork is perturbed.
Despite this knowledge, investigations of signaling networks and
regulating events have for a long time been recorded as averages for large
numbers of cells or molecules, often investigated in vitro. Unfortunately, it is
quite likely that such average analyses hide intercellular variations, and
analysis of individual cells will thus retrieve more complex, but also more
reliable information (Figure 1)5. In particular, as soon as we investigate not
cell cultures with clonal expansions from one precursor cell but mixtures of
cells, patient samples or tissue sections, averaging also means ignoring the
heterogeneity of different cell types within a tissue and what effects the
composition of this tissue has on each individual cell. In situ studies and
studies focusing on single cells have the advantage that these can be
correlated to the surrounding tissues. Thereby additional local,
morphological and, when molecules are investigated in vivo, in the living
cell or organism, maybe even spatiotemporal information can be retrieved.
11
Figure 1. Difference between single cell- and average detection methods.
All three cell populations have an average of 4 molecules per cell, which would be
the number detected by methods investigating the average of a sample. Studies with
single cell resolution would however reveal different degrees of cell-to-cell variation
in the left and right population and a truly homogenous distribution of signals in the
middle panel.
Heterogeneity is not only found between individual cells but also between
properties of individual molecules in these cells. In order to gain information
about distribution of these molecules in time and space, their interaction
partners and modifications, as well as about any changes upon different
stimuli, molecules have to be studied individually. Even if not all molecules
in a volume – for instance a cell – can be analyzed, single molecule analysis
will result in more than one (average) data point for the whole sample and
thereby strengthen the data6-8.
Of course, not all questions need to, can or should be answered by single
molecule / single cell methods. Very important findings were made with
averaging methods and could only have been found in that way. Still, in
order to comprehensively understand molecules, cells, tissues and
organisms, in health and disease, we have to move closer, look deeper and in
greater detail on what is happening to each molecule, in every cell in its
natural environment.
12
Methods to study protein-protein interactions in situ
Cell signaling is achieved by a series of connected events were proteinprotein interactions (PPIs) will add posttranslational modifications (PTMs)
to interacting proteins, causing conformational changes and activating the
modified protein for a subsequent PPI with another partner. This cascade of
PTMs and PPIs can transmit and amplify signals through the different
compartments of the cell, from the cellular borders to the nucleus and
specific genetic loci, and it is regulated by feedback loops and crosstalk
between pathways to ensure a controlled response to stimuli. Several
methods have been developed to study such interactions in vitro either for
screening purposes – e.g. 2D gel electrophoresis with mass spectrometry
readout9, yeast-two-hybrid systems10-11 or protein microarrays12-13 – or to
investigate certain interactions in greater detail by methods like coimmunoprecipitation14, protein fragment complementation assay (PCA)15
and the proximity ligation assay (PLA)16. These methods have helped to
unravel important information, however, with the exception of PLA they all
provide averages for populations of molecules or cells. As pointed out
above, in situ analyses are required in order to retrieve information on a
single cell level.
To study proteins in intact cells and tissues, methods to recognize and
label the proteins of interest are required, using either genetic modifications
to directly attach a reporter molecule to the targeted protein, or affinity
reagents (e.g. antibodies) to label the targeted protein with a reporter
molecule.
Fluorescence immunostaining (IF) is an example of the latter, applying
wide-field or confocal microscopy for readout. Interacting proteins are
detected in situ by two different primary antibodies. Either these antibodies
are directly labeled or they are in turn detected by fluorescence-labeled
secondary binders. If the proteins are co-localized, the fluorophores also
become co-localized in the cell (Figure 2). When the images derived from
the individual fluorescence channels are superimposed, the area of colocalization can be calculated by comparing fluorescence intensities in each
pixel of the image. Investigating PPI through confocal microscopy has the
benefit that spatial information about the PPI can be obtained at microscopic
resolution just below micrometers. In order to prevent detection of artifacts,
the aim is to visualize interactions of endogenously expressed proteins,
rather than overexpressed proteins, genetically modified with a fluorescence
tag. However, in some cases, low abundance of proteins can make detection
of few complexes challenging17.
13
Figure 2. Immunofluorescence (IF).
(A) Interacting proteins are detected by directly fluorescence-labeled antibodies or
(B) by primary antibodies derived from different species, which are in turn
visualized by fluorescence-labeled secondary antibodies.
The sensitivity of immunostaining is highly dependent on whether the signal
generated by binding of the affinity reagents, i.e. fluorescence-labeled
primary or secondary antibodies, can be distinguished from background
autofluorescence of the cells. Therefore, application of secondary
fluorescence-labeled antibodies instead of directly labeled primary
antibodies has the advantage of signal amplification: Every primary antibody
can be bound by several secondary antibodies (Figure 2B). This leads to an
increase of fluorophore concentration at the site where the primary antibody
has bound and hence a stronger signal.
To increase the spatial resolution of PPI detection, methods based on
resonance energy transfer can be used, reducing the distance required
between two proteins to record them as interacting molecules. Förster
resonance energy transfer (also: Fluorescence resonance energy transfer
(FRET)) exploits a nonradiative energy transfer between two adjacent
fluorophores with compatible excitation and emission spectra. Since the
emission spectrum of a fluorophore is of a longer wavelength span than its
excitation spectrum, a donor molecule can emit energy that will excite a
second fluorophore (acceptor) (Figure 3). This is however only possible if
the excitation spectrum is overlapping with the donors’ emission spectrum
and if the donors’ emission dipole and the acceptors’ excitation dipole are
not perpendicular orientated towards each other. Furthermore, nonradiative
energy transfer between donor and acceptor can only occur in a range of 210 nm. Beyond this distance, the energy has decayed too much to excite the
acceptor molecule18.
14
Figure 3. Förster resonance energy transfer (FRET).
(A) Two potentially interacting molecules are fused to two different fluorophores
with overlapping excitation- and emission spectra. When the proteins are not
interacting i.e. their distance from each other is larger than 10 nm, the donor is
excited and emits light at a longer wave length. (B) Upon interaction of the
molecules, the fluorophores are brought into FRET distance. If now the donor
fluorophore is excited, an energy transfer to the acceptor fluorophore occurs, leading
to (C) excitation of the acceptor which in turn starts to emit light.
Donor and acceptor molecules – fused to the proteins of interest – are
transfected into cells to be studied. When the target molecules are interacting
and thereby brought within FRET distance, energy transfer from the excited
donor to the acceptor occurs, leading to immediate excitation of the acceptor.
This energy transfer is detectable by different means, either by detecting the
increase of acceptor fluorescence intensity, changes in donor lifetime
(fluorescence lifetime imaging (FLIM)) or donor intensity upon addition or
photobleaching of the acceptor. Alternatively changes of the fluorophore
orientation towards each other can be detected18. By using FRET, the
dynamics of proximity between two proteins can be detected in situ and in
vivo19. Another possibility is to detect endogenous proteins with a pair of
antibodies labeled with the donor and acceptor molecules, respectively20.
Furthermore, protocols are available for single molecule detection in situ21,
as well as for detection of three interacting proteins22.
Bioluminescent resonance energy transfer (BRET) is closely related to
FRET, but employs a bioluminescent luciferase instead of the donor
fluorophore. Under oxidative conditions, the luciferase converts a substrate,
leading to the emission of light. If the acceptor fluorescent molecule is in
close proximity, then Förster resonance energy transfer will transport the
energy to the acceptor molecule and the fluorophore will be excited (Figure
15
4)23. BRET is able to circumvent some of the problems encountered with
FRET e.g. the risk for simultaneous excitation of both, donor and acceptor
fluorophores, due to overlapping excitation spectra. Furthermore, BRET
reactions do not require a certain orientation of the interacting proteins
towards each other since the energy that the luciferase sets free is spherically
distributed. The assay is not impaired by autofluorescence and does not
damage light sensitive tissue. BRET has recently been applied for detection
of PPI in living mice24 and at subcellular localization25.
Figure 4. Bioluminescent resonance energy transfer (BRET).
(A) One of the interacting proteins is fused to a luciferase, the other one to a
compatible fluorophore. The luciferase converts a substrate leading to the emission
of light as long as the proteins of interest are spatially separated by > 10 nm. (B)
When the proteins interact, the energy releasing enzyme and the fluorophore are
brought in proximity to each other and hence the energy is transferred to the
fluorophore, (C) excites it and light is emitted.
However, both FRET and BRET, by fusing the fluorophores / luciferase to
the proteins of interest, alter the proteins and their function may be
compromised. Hence careful controls are needed to prevent false-positive
and -negative signals. Expression levels have to be sufficiently high,
otherwise the energy transfer cannot be detected but not so high that the
physiological function of the proteins is impaired. For FRET, precautions
have also to be taken for the design of the fusion construct because the
orientation of donor and acceptor towards each other plays an important role
for FRET efficiency. Unfortunately, FRET is prone to photobleaching and
difficult to use in samples with high autofluorescence, limiting its usefulness
for studies in tissue sections. Furthermore, it may be challenging to find
fluorophores with excitation and emission spectra suited for higher
multiplexed FRET assays to simultaneously detect multiple interactions.
16
However, BRET also as drawbacks in that it requires a substrate which has
to be delivered into the cell. Furthermore, the substrate is not locally
anchored, and it is therefore difficult to create a localized signal23.
PCA is another approach that can be used to detect PPIs, based upon
reassembly of a split reporter molecule to obtain a functional reporter
molecule. The approach was originally described using ubiquitin as the
reporter molecule15. Bimolecular fluorescence complementation (BiFC) is a
special case of PCA, based on the assembly of a fluorescent protein from
two non-fluorescent domains, each domain fused to one protein of a
complex. Upon complex formation of these proteins the fluorescent protein
is reconstituted from both domains, hence fluorescence from non-interacting
molecules is avoided (Figure 5)26. This property renders BiFC more sensitive
than F/BRET26-27. When the two proteins to be investigated interact in a
living cell, the location of the association can be visualized in situ and even
weak and transient interactions can be detected as the interactions are
preserved by the assembled fluorophores. Interaction partners do not need to
be positioned in a certain orientation towards each other as it is the case with
FRET, and the distance range for BiFC depends on the length of the linkers.
An advantage of BiFC over other forms of PCA is that it does not depend on
exogenous substrates28. BiFC can be utilized to study PPIs in living cells29
and has been used for interaction studies in several organisms, e.g.
mammalian cells, plants30 and yeast31. Multiplexed detection is enabled by
using several pairs of split fluorescent molecules32.
The fact that BiFC suffers from a delay of one hour or more between
complex formation of the proteins and reconstitution of the fluorescent
molecule is however a drawback. Moreover, dissociation detection is
difficult due to the stability of the reconstituted protein domains. These two
properties severely limit opportunities to use the method for studies on
dynamic events26, and the stable complexes may interfere with cellular
physiology.
Since it might be challenging to multiplex each method in itself, detection
of higher order complex formation is accomplished through combinations of
the methods described above: BiFC and FRET33, BiFC and BRET34 or FRET
and BRET35.
17
Figure 5. Bimolecular fluorescence energy transfer (BiFC).
(A) A fluorescent molecule is split into two halves that are fused to the proteins of
interest. (B) Upon interaction of the target proteins the halves are brought in
proximity to each other, and can reconstitute to form a fluorescent molecule (bulb),
(C) which can subsequently be excited to emit light.
The methods discussed so far for in situ detection of PPI share the
limitations that they often require the studied proteins to be fused to domains
and transfected into cells, potentially leading to unphysiological
overexpression. The fluorescence signal generated may be difficult to detect
in cells and tissue sections with high autofluorescence. To overcome these
limitations, namely to detect endogenous PPIs in fixed cell lines and tissues,
the in situ proximity ligation assay (in situ PLA) was developed (Figure 6)36.
This method is based on the proximity ligation assay (PLA)16, where a
double recognition of the protein of interest through oligonucleotidecarrying antibodies (PLA-probes)37 or aptamers16 is required to produce an
amplifiable reporter molecule. Upon proximal binding of the PLA-probes,
the oligonucleotides can be joined by ligation and amplified by a PCR
reaction.
However, to allow localized detection of the protein complexes in situ,
another method for DNA amplification is required, i.e. rolling circle
amplification (RCA). For in situ PLA the oligonucleotides on the PLAprobes instead act as a template for hybridization and ligation of two
subsequently added connector oligonucleotides, which can form a circular
DNA molecule. Once ligated, this circular molecule can be amplified by
RCA, primed from the oligonucleotide attached to the PLA-probe. After the
first round of amplification, phi29 DNA polymerase – a polymerase with a
strong strand displacement activity – displaces the newly created strand
18
leading to the production of single-stranded concatemeric complements of
the original DNA circle38-39. As the RCA-priming molecule stays attached to
the detected protein complex via the antibody, the amplified DNAconcatemer will remain located at the site where the protein complex was
detected. Such immobilized RCA-products (RCPs) are visualized by
hybridization of fluorescence-labeled detection oligonucleotides binding to a
hybridization sequence encoded in the DNA-circle. Since the amplified
product consists of hundreds of copies, hundreds of detection
oligonucleotides will hybridize to the same molecule. The resulting bright
signals, approximately one µm in size, are easily distinguishable from
potential background fluorescence in an epifluorescence microscope. Each
detected interaction event gives rise to one distinct RCP and it is therefore
straightforward to enumerate detected complexes. In situ PLA provides an
assay for detection of endogenous proteins and PPIs in cells and tissue
sections, with a single-molecule resolution and the possibility for relative
quantification of the signal. Further developments of the method include the
development of a brightfield variant40, application of secondary PLA-probes
for detection of PTMs (paper I), flow cytometry as readout platform (paper
II) and for detection of three proteins forming one complex36.
19
Figure 6. In situ proximity ligation assay (in situ PLA).
(A) Two PLA-probes (red, blue), consisting of an antibody and an oligonucleotide
each, can bind to the protein of interest. (B) Upon complex formation, the PLAprobes are brought in proximity. Now two connector oligonucleotides can hybridize
and (C) be ligated forming a circular molecule. (D) This DNA circle can now be
amplified as a concatemeric copy of the circle by RCA and be detected with
fluorescence-labeled detection oligonucleotides (stars). (E) Each resulting RCP and
with that each detected interaction or modification (red dots, see green rectangle) is
visualized as single bright spots in a fluorescence microscope. Here phosphorylated
PDGFR in BJhTert cells detected by in situ PLA with secondary PLA-probes is
depicted.
Methods to study protein-DNA interactions in situ
Although all cells in the body contain the same genomic DNA, the
accessibility of this information for the cell is closely regulated by proteins,
e.g. the DNA is tightly wrapped around histones organizing the DNA in
nucleosomes. It is also proteins that direct transcription of the genes for
expression (transcription factors and RNA polymerases) and copy DNA
prior to cell division (DNA polymerases and ligases). For all such and many
more events, proteins are interacting with nucleic acids. Therefore, I think
that it is as important to study protein-DNA interactions (PDI) and histone
modifications, how access of DNA is regulated and what impact that might
have on a cell’s fate in situ, as it is important to study protein and mRNA
20
expression, PPIs and protein modifications. However, today there is no
method available to study individual PDIs in single eukaryotic cells and it
can therefore only be hypothesized that there is an equally large variation
between individual cells within a seemingly homogenous population as it
has been seen for proteins and mRNA-molecules41-43.
There are several methods that can extract information about proteins
binding to DNA-sequences, one of the most popular ones being chromatin
immunoprecipitation (ChIP)44-45. ChIP extracts DNA-sequences interacting
with a protein by applying an antibody against the protein of interest to
fragmented nuclear extract. By doing so, the DNA fragments bound to the
protein will be retrieved together with the protein. After reversal
crosslinking, DNA bound by the protein can be determined by PCR,
microarrays (ChIP-on-chip) or sequencing (ChIP-seq)46-47. With ChIP, new
interactions can be found but they are averaged over hundred48 to millions of
cells and no spatial information can be retrieved, rendering the method
unsuited to study cell-to-cell variations.
PLA has also been applied for detection of PDIs. The binding of
recombinant transcription factors to synthetic oligonucleotides spiked into
nuclear extracts has been investigated. By using a panel of synthetic
oligonucleotide sequences containing potential binding sites for the
transcription factor, a consensus sequence for the transcription factor
recognition site could be determined. As described above, one PLA-probe
was used to detect the protein. Instead of the second PLA-probe, a part of the
target DNA-sequence was allowed to participate in the ligation reaction.
Upon binding of the protein to the right DNA-sequence, the oligonucleotide
on the PLA-probe could be ligated to the bound DNA with the help of a
connector oligonucleotide. Detection of interactions of the amount of
proteins corresponding to 1-10 cells was analyzed using qPCR-readout49.
Nevertheless, none of the methods described so far can be used to retrieve
spatial information from individual cells – in which cells do these
interactions or epigenetic modifications occur and how often and where in
the tissue? In order to answer questions related to variation in the population,
in situ methods are required. The major obstacle for such analyses is to find
a compromise between making the DNA-sequence accessible and at the
same time keeping proteins bound to the DNA. Furthermore, detection of a
specific DNA-sequence can be challenging, especially if this sequence is
short. For detection of PDIs in situ FRET based assays have been used. In
one study the proteins to be analyzed were fused to green fluorescent protein
(GFP), serving as donor fluorophore, and transfected into mammalian cells.
A fluorescent dye, intercalating with genomic DNA, was used as acceptor
(Figure 7). Upon close proximity between the investigated protein (e.g. the
glucocorticoid receptor and heterochromatin proteins 1 and 1 ) and the
genomic DNA, energy is transferred from the excited GFP fused protein to
the DNA intercalating dye and changes in donor lifetime can be measured by
21
FRET-FLIM50. Even though this method allows investigation of differences
between individual cells, no information about the DNA-sequence can be
retrieved.
Figure 7. FRET-FLIM.
(A) Genomic DNA is
labeled with a fluorescent
dye suitable as FRET
acceptor (dark grey stars)
and
DNA-binding
proteins are labeled with
the corresponding FRET
donor.
(B)
The
fluorophore on the protein
can be excited and upon
binding of the protein to
the DNA FRET occurs
between
fluorescent
molecules within FRET
distance. (C) This leads to
excitation of the acceptors
which in turn start to emit
light.
To detect a specific DNA-sequence and its co-localization with a particular
protein, IF and fluorescence in situ hybridization were combined to immuno
DNA fluorescence in situ hybridization (immuno DNA FISH) and immuno
RNA FISH 51-53. First, antibodies against the protein of interest are applied,
followed by secondary fluorescence-labeled antibodies. The cells are then
postfixated before the genomic DNA is denatured and the target sequence
detected by hybridization of hapten labeled detection oligonucleotides.
22
These haptens can in turn be detected by IF utilizing antibodies against the
label (Figure 8). By superimposing the different fluorescent channels, colocalization of the nucleic acid sequence and the protein can be detected and
visualized in situ. Immuno DNA FISH has e.g. been applied for colocalization studies of histone methylations in chromosomal subdomains and
at differentially expressed gene loci in human cells54. Although immuno
DNA / RNA FISH presents a valuable tool for investigation of PDIs in situ,
it is still limited in resolution since the target nucleic acid sequence is
recognized by hybridization and has to be quite large (see below).
Furthermore, spatial resolution is restricted to the resolution of the
microscope used and proteins can only be detected if they are so abundant
that IF can distinguish them from background.
Figure 8. Immuno DNA / RNA FISH.
Upon fixation, the protein of interest is detected by IF and the secondary antibodies
are labeled with one color (red stars). Subsequently, the DNA-sequence is detected
with several labeled probes (green sequence), which in turn are detected by IF
(green stars).
23
Methods to study nucleic acids in situ
To my knowledge there is today no method available for in situ PDI
detection with a sequence resolution better than several kbs to Mbs.
However, methods for detection of plain nucleic acid sequences that can
discriminate between shorter sequences exist:
In in situ hybridization (ISH) the target DNA-sequence is detected by
hybridization, as described for immuno DNA FISH, with radioactively
labeled probes55-56, but this detection is limited in resolution and by the probe
instability. Therefore ISH was further developed, utilizing other means to
label the detection probe e.g. enzymatic reactions and fluorescence-labeled
probes (FISH), testing different probe lengths and extent of probe labeling.
Using either directly fluorescence-labeled or hapten-labeled probes
amplified by secondary detection, signals from hybridized sequences could
be visualized: several Mbs large DNA-sequences in metaphase spreads and
down to ~50 kb in interphase chromatin57. On the other end of the scale,
larger sequences up to whole chromosomes – even in 24-plex – and
individual chromosome arms in 48-plex could be simultaneously and
differentially labeled58-60.
Also, detection of single mRNA molecules by multiple fluorescencelabeled probes was accomplished. Single mRNA molecules were either
detected by few, multiply labeled, or by several, singly labeled
oligonucleotides43,61-62. Incorporation of the RNA analogous locked nucleic
acid (LNA) into the probe was utilized to increase hybridization affinity.
Hence higher hybridization temperatures could be applied and with that
higher sequence discrimination accomplished. Utilizing LNA-containing
probes, miRNAs could be detected in situ applying signal amplification
schemes like tyramide signal amplification (TSA) or enzyme-labeled
fluorescence. With the latter approach even individual miRNA molecules
could be visualized63-64. LNA as well as peptide nucleic acid – DNA analogs
which bind stronger to DNA due to their uncharged peptide backbone –
containing probes form more stable complexes with the target DNA and
enhance thereby detection sensitivity in FISH65. Furthermore, FISH has been
applied for detection of DNA and mRNA-sequences in vivo66-67.
Primed in situ labeling (PRINS)68 is mostly applied for detection of
repeated sequences and utilizes a specific oligonucleotide to hybridize to the
target sequence in denatured chromosomes. Upon hybridization, the primer
can be elongated by a DNA polymerase, incorporating hapten labeled
nucleotides. These nucleotides can be detected by fluorescence-labeled
avidin molecules or anti-hapten antibodies allowing visualization of the
target sequence in situ. Even though single nucleotide polymorphism (SNP)discrimination can be accomplished when positioned at the 3’ end of the
primer, detection of individual genes is difficult, probably since the created
signal is not strong enough and not reaching high enough detection
24
efficiency – impaired by gaps in the target DNA, which can serve as
undesired primers. However, single gene detection has been accomplished
combining PRINS with signal amplification by TSA69-70.
In situ PCR and in situ reverse transcriptase PCR amplify detected DNA
or cDNA-sequences exponentially, thereby increasing their availability for
FISH detection, which leads to a decrease in the limit of detection. Detection
is utilized either directly by fluorescence labeling the PCR product or
indirectly utilizing haptens. Since the signal is amplified, shorter sequences
(150-350 bp) can be visualized and in situ PCR was successfully applied to
single copy genes and low copy RNAs. Although the resulting PCR-products
are fixed to the cells, they are not anchored at the site where they were
generated. PCR-products can therefore diffuse away, causing false positive
signals while false negative signals occur due to inefficient in situ PCR
reactions71-72.
Another approach to amplify the detected target sequence is by the
branched DNA technique73. The target DNA-sequence is detected by several
oligonucleotides hybridizing only at the 3’ end to the target sequence. The
free 5’ end can in turn be bound by a preamplifier sequence consisting of
tandem repeats, which are bound by several amplifier oligonucleotides. This
structure can be detected by alkaline phosphatase labeled probes directed
against the preamplifier creating a comb-like structure which strongly
amplifies the localized signal.
Padlock probes – a technique capable of distinguishing single molecule
sequence variants in situ – is a further development of the oligonucleotide
ligation assay (OLA)74. In OLA, two oligonucleotides that hybridize
adjacent to each other are ligated by T4-DNA ligase if they form a perfect
match with the target sequence. For padlock probes these two hybridization
oligonucleotides are linked together by a spacer region. Hybridization to its
target sequence brings the 3’ and 5’ ends together, facing each other75-76.
Only a perfect match in the junction enables ligation, which creates a
circular DNA molecule that subsequently can be detected. For in situ
detection of single molecules however detection by hapten- and
fluorescence-labeled probes was not sensitive enough. Therefore, padlock
probes were combined with efficient and locally anchored signal
amplification by RCA, primed from enzymatically cleaved target DNA77.
The RCP can then be detected by hybridization of fluorescence-labeled
detection oligonucleotides (Figure 9). Padlock probes have been applied for
detection of mitochondrial DNA in fixed cells in situ77 and genomic DNA in
comet assays78 as well as recently for detection of mRNA in situ79. For this
application the mRNA molecules are first converted to target cDNA through
reverse transcription using specific LNA-containing primers. mRNA is then
degraded with RNase H to provide a template for padlock hybridization,
except for the part where the primer is bound, since the LNA bases protect
the mRNA from degradation. This ensures that the cDNA stays attached to
25
the place where the mRNA was located and provides a cDNA template for
the padlock probes, as ligation of DNA probes hybridized to RNA is less
efficient. Padlock probes have been applied for SNP discrimination in situ
and multiplexed mRNA detection, enabled by the linker sequence including
the detection sequence, which can be freely chosen79.
Figure 9. Padlock probe.
(A) A padlock probe is a
single
stranded
DNA
molecule with the 3’ and 5’
ends
(red,
blue)
complementary to the target
strand. A segment of the
noncomplementary part of
the probe is utilized as
detection site (green). (B)
Upon hybridization of the
probe to its target sequence,
the 3’ and 5’ ends of the
probe are brought together
and can be ligated to a
circular
DNA-molecule.
(C) Now an RCA can be
primed from the target
molecule and the resulting
product can be detected by
a
fluorescence-labeled
detection oligonucleotide
(green sequence with red
stars).
26
Model systems
PDGFR and DUSP6
Platelet-derived growth factor receptor beta (PDGFR belongs to a family
of receptor protein-tyrosine kinases (RTKs), characterized by an
extracellular, ligand binding domain composed of five immunoglobulin-like
loops, a transmembrane domain, a juxtamembrane domain, two tyrosine
kinase domains separated by a kinase insert domain and a C-terminal
domain. Stimulation with platelet-derived growth factor-BB (PDGF-BB)80 or
-DD81 leads to dimerization of two receptor molecules, conformational
changes and autophosphorylation at multiple tyrosine residues in the
intracellular domains of the receptors. Thereby binding sites for SH2-domain
containing downstream signaling mediators are created e.g. the RAS-RAFMEK-ERK-pathway mediating an increase in cell survival, proliferation and
motility82-83. Upon sustained signal activation the receptor is downregulated
by internalization of the ligand-receptor complex. Dysregulated PDGFR
signaling is found in malignancies like dermatofibrosarcoma protuberance or
glioblastoma multiforme82 and the receptor is therefore a potential target for
treatment84. However, detecting the PDGFR specifically is challenging,
since the PDGFR shares ~80% sequence homology in the kinase domain
with the alpha isoform of the PDGFR85 and both isoforms are equally well
stimulated by PDGF-BB86. This has rendered it difficult to distinguish
between PDGFR and PDGFR especially in their activated forms, using
methods like IF and immunohistochemistry (IHC).
Dual-specificity phosphatase 6 (DUSP6, alternatively called MKP-3) is an
ERK-specific cytosolic member of the MAP kinase phosphatases (MKP),
exhibiting phosphatase activity specific against threonine and tyrosine
residues in the activation loop of MAPK87. ERK is activated by RTK
signaling (including PDGFR 88) among others via GRB2 and SOS and the
RAS-RAF-MEK-pathway. Upon activation, phosphorylated ERK
translocates into the nucleus and activates transcription of target genes. One
of these targets is DUSP6, which in turn becomes expressed,
dephosphorylates ERK and therefore forms a feed-back-loop for ERK
activation.
27
EGFR family
The epidermal growth factor receptor (EGFR) belongs to the family of
receptor tyrosine kinases; consisting of four members: EGFR and HER2-4.
All members comprise an extracellular binding domain, a lipophilic
transmembrane segment, and an intracellular tyrosine kinase domain, which
is inactive in HER3. Through stimulation with different ligands, the EGFR
family members can form homo- and heterodimers. Binding of six different
ligands to EGFR has been described, among them EGF and TNF- , while no
ligands have yet been reported for HER2. Upon stimulation, EGFR is
phosphorylated at tyrosine residues in the intracellular domain, leading to
recruitment of effector proteins and signal propagation in the cell, via RASRAF-MEK- and PI3K- pathways89. Whether the receptor dimerizes upon
stimulation90 or exists as preformed dimers/complexes that undergo a
conformational change upon ligand binding, leading to phosphorylation of
the intracellular domains91, is however still a controversy. EGFR and HER2
are both overexpressed in many human cancers – breast, ovarian and
stomach cancer to name a few – and are therefore possible targets for cancer
therapy. Monoclonal antibodies against HER2 – Trastuzumab – and EGFR –
Cetuximab – are currently applied in the clinic, i.e. Trastuzumab was
approved for treatment of HER2 overexpressing metastatic breast cancer and
Cetuximab for treatment of colorectal cancer as well as head and neck
cancer.
Chromatin structure and Alu-repeats
In its smallest entity, chromatin is organized in nucleosomes, each consisting
of an octamer histone core of two halves built from histone H2A, H2B, H3
and H4. Around this core, genomic DNA is wound in two turns consisting of
147 bp. Where the DNA enters and leaves the histone core, it is retained by
one of two histone molecules, either H1 or H5, which compact the DNA
further into a fiber structure (Figure 10). Whether a certain DNA-sequence is
accessible for transcription is regulated by epigenetic signals acting either in
trans – through feedback loops of transcription factors back on the gene
itself and RNA-directed positioning of epigenetic signals – or cis –
modifications, which are stably associated with the DNA-sequence and
inherited with it92. Such cis-acting modifications are e.g. methylation of
DNA in CpG islands and histone modifications such as methylation,
acetylation, ubiquitylation and phosphorylation to name a few. Certain
histone modifications are associated with specific effects on the
transcriptional activity. For example, H3K9 methylation is associated with
transcriptional inactivity, while H3K4 methylation is associated with active
transcription.
28
Figure 10. Chromatin structure.
(A) Genomic DNA is wrapped
around a histone core consisting of 8
histones (two H2A, H2B, H3 and H4
respectively, dark grey circles).
Histone H1or H5 is utilized as linker
histone (light grey). (B) The
nucleosomes are then organized like
“beads-on-a-string”
and
(C)
condensed into a ~30 nm fiber. (D)
This fiber is in turn abridged into
active chromatin, which (E) can be
further condensed to a chromosome.
The family of Alu repeats belongs to the family of short interspersed repeats
(SINE) present in over 1 million times in the human genome93, being one of
the larger families of the SINE accounting for ~5% of human DNA94. The
family is divided in several subfamilies with AluJ and AluS being the oldest
and most abundant ones and AluY, which is still active as transposable
element. Most of the Alu-repeats can be cleaved by AluI restriction
29
endonuclease. The primate sequence consists of ~300 bp, two imperfect
repeated monomers, whereof one is well conserved between different Alufamily members and the other monomer being more variable. The rodent
sequence contains only one monomer of ~130 bp. In 1995, a short 26 bp Alu
consensus sequence was identified95, which is also present in the left arm of
the ~300 bp Alu consensus sequence presented by Deidinger et al.96. This
short consensus sequence is present ~60,000 times in the human genome
(perfect matches counted in NCBI, assembly march 2006). Alu repeats are
not randomly distributed in the genome but rather seem to accumulate in
gene rich regions93. Alu elements are mobile sequences, their genome
interactions are suggested to be recombinatorial hotspots and involved
through homologous recombination in most gene rearrangements95. Alu
repeats can cause diseases by creating new insertions into coding sequences
or promoters and by altering the level of gene expression, changing the
reading frame and causing deletions and duplications by unequal cross
over94. Furthermore, Alu repeats were found to be hypomethylated in colon
cancer cell lines when compared to normal colon epithelial cells and the
CpG island methylation level at one specific AluY repeat correlated with the
amount of trimethylated histone H3K9 in these cell lines97. Hence analysis of
the methylation status of repetitive elements at specific genomic loci could
provide new insights into the mechanisms for cancer development.
30
Present investigations
The papers included in this thesis are all based on the development of novel
methods for in situ detection of individual PPIs and protein modifications by
in situ PLA, either alone (paper I and II) or together with simultaneous
detection of individual mRNA-molecules by padlock probes (paper III). In
paper IV, in situ PLA and padlock probes were combined as a novel method
for detection of individual, sequence specific protein-DNA interactions.
Paper I: In situ detection of phosphorylated plateletderived growth factor receptor using a generalized
proximity ligation method
Aim of the study
In situ PLA36 was generalized by exploiting secondary, species-specific
PLA-probes directed against primary antibodies derived from different
species. These probes were applied for detection of individual
phosphorylated PDGFR in transfected as well as endogenously expressing
cells and tissue sections. Earlier, the phosphorylation status of the PDGFR
has been investigated in situ using IHC or IF with phospho-specific
antibodies against PDGFR 98-99. Due to high sequence homology in the
kinase domains85, it has been difficult to distinguish phosphorylated
PDGFR from the alpha-isoform. By combining the phospho-tyrosinespecific antibody with a receptor-specific antibody, requiring the binding of
two antibodies to generate a signal with in situ PLA, we gained the
selectivity that allowed us to distinguish both isoforms and a signal
amplification to visualize individual receptors. An additional value by using
secondary PLA-probes was that they could be conjugated in larger batches,
and the need to conjugate every primary antibody was thus circumvented.
Methods
All cell types utilized in this paper – HEK293-PDGFR , PAE cells
overexpressing either PDGFR (PAE ) or PDGFR (PAE ) and human
fibroblasts (BJhTert) endogenously expressing the receptor – were starved
31
for 24 h prior to stimulation with PDGF-BB at indicated concentrations.
After fixation and blocking, primary antibodies rabbit-anti-PDGFR and
mouse-anti-PDGFR -pY751 were applied. Upon binding of the primary
antibodies to the same target protein complex, secondary PLA-probes
directed against mouse- or rabbit-IgG could bind to their target proteins
thereby bringing the oligonucleotides in close proximity to each other.
Subsequently, the assay was performed as described above for the original in
situ PLA36, and RCPs representing single phosphorylated PDGFR molecules were quantified by image analysis. Detection of phosphorylated
PDGFR in fresh frozen tissue sections was similar to detection in fixed
cells except that higher concentrations of primary antibodies, PLA-probes
and phi29-DNA polymerase were used.
For immunoblot analysis of phosphorylated PDGFR , the cells were
starved and stimulated with different concentrations of PDGF-BB before
they were lyzed and subjected to SDS-PAGE and immunoblotting. The
membrane was probed with rabbit-anti-PDGFR antibody and subsequently
stripped and reprobed with mouse-anti-pY751 antibody.
Results
Phosphorylation of PDGFR was first investigated in HEK293 cells stably
over-expressing PDGFR by in situ PLA. In unstimulated cells, minimal
amounts of signals were detectable. Upon stimulation with PDGF-BB, the
amount of signals increased dramatically. As expected, technical controls
confirmed, that presence of the phosphorylated receptor, primary antibodies
as well as both PLA-probes were required for RCPs to arise.
Selectivity of in situ PLA was demonstrated using stably transfected
PAE-cells expressing either the alpha- or beta-isoform of PDGFR. Both
isoforms are stimulated by PDGF-BB to the same extent and closely
related85-86. Using in situ PLA we detected PDGFR phosphorylation in the
beta-isoform expressing cells, while only negligible amounts were found in
the alpha-isoform expressing cells.
It was possible to detect phosphorylated PDGFR even in endogenously
expressing human fibroblasts. When stimulating the cells with different
amounts of PDGF-BB, in situ PLA revealed a similar dose-response curve as
obtained with immunoblotting.
To test if the assay would be applicable for analyses in tissue sections,
fresh frozen human scar tissue was analyzed for the presence of
phosphorylated PDGFR Receptor expressing stromal fibroblasts and
pericytes are stimulated by PDGF-BB secreted during wound healing100, so
in situ PLA signals were found in the fibrotic dermal stroma beneath the
epithelium and the stroma around venules, in accordance with literature.
32
Discussion
In situ PLA was shown to be a potent method also for detection of PTMs.
Double recognition of the target protein for signal generation and the
therewith accomplished selectivity, made it possible to discriminate between
the phosphorylated PDGFR and - The potent amplification step
integrated in the protocol enabled detection of single phosphorylated
receptors. By utilizing species-specific secondary PLA-probes, the need to
conjugate every primary antibody was overcome. Thus, secondary PLAprobes can be produced in large batches and be stored for a long time,
ensuring reproducibility of the experiments. Any pair of primary antibodies
can now be applied as long as the two antibodies are derived from two
different species.
Since the RCPs stay attached to the site where the primary antibodies
have bound, the results can hence be interpreted both with regard to the
spatial distribution of the signals within the cells, as well as the variation
within the cell population.
Paper II: Flow cytometric in situ proximity ligation
analyses of protein interactions and post-translational
modification of the epidermal growth factor receptor
family
Aim of the study
In paper II, in situ PLA was performed using flow cytometry as readout,
which would enable analysis of thousands of individual cells in one run. In
this study, homo- and heterodimers between EGFR and HER2 were
determined and detection of the phosphorylation status of EGFR by in situ
PLA was compared to IF.
Methods
For relative quantification of EGFR and HER2; SKOV-3, U343 and MCF-7
cells were stained with fluorescence conjugated primary antibodies and
analyzed using a flow cytometer. The same cells were analyzed with in situ
PLA for EGFR and HER2 interactions using PLA-probes made from
Cetuximab and Trastuzumab. To visualize phosphorylated EGFR in EGFstimulated U343 by in situ PLA, primary antibodies goat-phosphorylatedEGFR (pEGFR, pY1173) and rabbit-pEGFR (pY1045) together with
secondary PLA-probes were applied. The same primary antibody (rabbitpEGFR pY1045) was also used in IF and both samples were analyzed by
flow cytometry.
33
For comparison, cytospins of the same EGF-stimulated cells were used to
detect phosphorylated EGFR either by IF using the rabbit-anti-pEGFR
antibody or by in situ PLA.
Results
Levels of homo- and heterodimers of EGFR and HER2 were investigated in
three different cell lines: U343, SKOV-3 and MCF-7. First the protein levels
were determined by IF-based flow cytometry: U343 and SKOV-3 expressed
high levels of EGFR while MCF-7 expressed only low levels; for HER2,
U343 expressed low, SKOV-3 high and MCF-7 intermediate levels of
protein. Next, in situ PLA-based flow cytometry was applied to detect levels
of EGFR-EGFR, EGFR-HER2 and HER2-HER2 dimers: U343 showed high
levels of EGFR-EGFR homodimers and low levels of the hetero- and HER2
homodimers. In SKOV-3 cells low levels of EGFR-EGFR but high levels of
the other two investigated dimers were detected. Finally, for MCF-7 none of
the investigated dimers could be detected even though HER2 was expressed
in these cells. However, MCF-7 also expresses HER3, which is a preferred
interaction partner of HER2 and might bind most of the HER2 molecules.
In the next step, to detect changes in dimerization upon stimulation with
EGF, the three cell lines were stimulated with varying amounts of EGF for
different amounts of time and at different temperatures. Since no changes in
dimerization were detected in any of these conditions, the phosphorylation
status of the EGFR in U343 cells upon stimulation by EGF was instead
determined by both in situ PLA and IF. When utilizing one of the primary
anti-pEGFR antibodies for IF to detect pEGFR, the signal increased only
slightly over background upon stimulation. With in situ PLA utilizing both
pEGFR-antibodies, the ligand dependent increase in signal was dramatic.
The same observation could be made when these experiments were
performed on cytospins of the stimulated cells.
Discussion
In our work we were able to detect homo- and heterodimers of EGFR and
HER2 by in situ PLA, and the results for U343 and SKOV-3 are in
agreement with what is described in the literature, namely that the preferred
interaction partner is HER2. In the MCF-7 cell line, we could not detect any
of these dimers, even though HER2 is expressed in this cell line but most
probably engaged with HER3. Therefore, expression levels alone do not
provide enough insight into how the proteins interact with each other. We
did not detect any changes in dimerization levels in any of the investigated
dimers upon EGF-stimulation, but could detect changes in the amount of
pEGFR by in situ PLA, confirming that the EGF-stimulation was successful.
This phosphorylation was difficult to detect by IF, due to unspecific staining
34
of the primary antibody, again showing that dual recognition provides
advantages over single recognition. The lack of response in dimerization,
upon ligand stimulation, might be due to that the distance requirement of our
in situ PLA (estimated to be below 30 nm) is not sufficient to detect the
change in distance between the epitopes of the EGFR, although for other
receptors such as VEGFR101 receptor dimerization upon stimulation has been
detected. Whether the members of the EGFR family exist as pre-formed
clusters in the cell membrane that change conformation upon stimulation
with the ligand, or whether they exists as inactive monomers, which
associate with each other upon stimulation, is widely discussed in the
literature90-91.
In this paper we have successfully used in situ PLA together with flow
cytometry, which increases the through-put compared to in situ assays by
microscopy, allowing investigation of thousands of individual cells. We
could also show that in situ PLA can be used for detection of PTMs even if
each antibody alone, when used in IF, is not selective enough to distinguish
its target molecule from the background.
Paper III: Simultaneous visualization of both signaling
cascade activity and end-point gene expression in single
cells.
Aim of the study
Two methods recently developed by my group were combined for detection
of PTMs of receptors (in situ PLA) and individual mRNA molecules
(padlock probes)79. In situ PLA was utilized for the detection of
phosphorylated PDGFR together with padlock probes for detection of
DUSP6 mRNA, a downstream target of the PDGFR signaling. The aim of
the study was to determine if the two protocols could be combined, since the
ability to address several nodes of a pathway at the same time would be a
valuable tool for e.g. drug screening.
Methods
BJhTert cells were starved for 48 hours in Modified Eagle Medium + 0.1%
fetal calf serum. For stimulation, PDGF-BB was added at different time
points and the cells were incubated at 37oC. When inhibitors (Gleevec or 5Iodotubercidin) were applied, these were added at a final concentration of 10
µM 1 h before stimulation with PDGF-BB was started. The drug
concentration was kept constant throughout the stimulation. After fixation
and permeabilization, mRNA detection by padlock probes was performed as
35
described in Larsson et al.79. Alternatively, PDGFR -phosphorylation was
detected as described in paper II, with the exception that the pY751-antibody
was exchanged for the pY100-antibody.
To accomplish simultaneous detection of phosphorylated PDGFR and
DUSP6 molecules, padlock probes were ligated to the target cDNA before
the cells were blocked and primary antibodies and PLA-probes were applied.
After hybridization and ligation of the circularization oligonucleotides for in
situ PLA, RCA was performed for both methods together and detected by
hybridization of differently labeled oligonucleotides. In all cases, the cell
membranes were stained with FITC-labeled wheat germ agglutinin and the
nuclei were labeled with DAPI. Images were taken with an epifluorescence
microscope, three images per condition, enhanced in ImageJ102 and analyzed
in CellProfiler103 on a single cell level.
Results
First the kinetics of PDGF-BB stimulation was determined for PDGFR
phosphorylation and DUSP6 expression levels. As expected, PDGFR
phosphorylation increased steeply within the first 5-10 min of stimulation
before the levels gradually decreased towards baseline levels, due to
internalization of the receptor. Maximal DUSP6 expression was detected
after 3 hours of PDGF-BB stimulation, in the same range as described
earlier, despite differences in the applied model system88. Then in situ PLA
and padlock probe detection were combined in one protocol. To access what
effect the combination of these two methods had, we investigated ACTB
mRNA instead of DUSP6 since ACTB is an abundantly expressed
housekeeping gene and changes in detection efficiency would be more
readily detectable. Even though the combining of both protocols decreased
the detection efficiency by 50% compared to the individual detections, the
levels of ACTB were overall unaffected by treatment with PDGF-BB,
indicating that a higher number of in situ PLA signals does not inhibit
detection by padlock probes.
Then, in situ PLA and padlock probes were combined to detect, which
effects the treatment with either Gleevec or 5-Iodotubercidin had on the
PDGFR pathway at the two nodes investigated. As expected, treatment
with Gleevec prior to PDGF-BB abolished the increase in signals from both
PDGFR phosphorylation and DUSP6 expression. 5-Iodotubercidin on the
other hand affected only signals from DUSP6, since 5-Iodotubercidin acts on
the pathway downstream of PDGFR .
36
Discussion
The results presented herein show that in situ PLA can be combined with
padlock probes for detection of individual proteins and mRNA molecules in
situ.
Since data analysis was done on a single cell level, cell-to-cell differences
were readily accessible. Even though the cells were synchronized by
starvation, the variation in response to PDGF-BB treatment was quite large
at the investigated time points in particular for DUSP6, where some cells
seemed to burst with expression while others were lacking behind.
Alternatively, transient bursts of expression in between the time points of
analysis were not detected. This heterogeneity would have gone unseen with
an averaging method. Furthermore, as the molecules are detected in situ, not
only information about the amount of detected signals per cell can be
retrieved, but these data can even be put in context of the morphology and
surrounding of the cell. The method presented here for simultaneous
detection of individual PPIs, PTMs of proteins and mRNA molecules should
be particularly interesting in drug screening applications for discovery of
new drugs that affect a pathway at a certain point but leave other parts
unaffected.
Paper IV: Visualizing individual, sequence-specific
protein-DNA interactions in situ
Aim of the study
The aim of this paper was to develop a method for detection of protein-DNA
interactions in situ that can discriminate between closely related genomic
sequences. As a model system, histone H3’s binding to a 26 bp-Alu repeat
consensus sequence (5’CCTGTAATCCCAGCACTTTGGGAGGC 3’)95
was utilized. This Alu-repeat sequence is highly abundant in humans
(~60,000 copies per genome) but not present in mice, although they carry an
orthologous
sequence
present
in
~6,000
copies
(5’CCTTTAATCCCAGCACTCGGGAGGC 3’; differences to the human
sequence are indicated in bold)104.
Three different approaches for detection of protein-DNA interactions in
situ were investigated. The most selective – “padlock probe based in situ
PLA” – based protein recognition on in situ PLA with a single antibody and
a single PLA-probe and detection of the genomic sequence on modified
padlock probes.
37
Methods
Human (BJhTert) and mouse (NIH3T3) fibroblasts were seeded on glass
slides and allowed to attach over night. Subsequent to fixation and
permeabilization, the cells were treated with AluI and Lambda-exonuclease
to cut the genomic DNA and make it partially single stranded. Now the cells
were subjected to one of the following protocols:
Hybridization based in situ PLA
As DNA-detecting PLA-probe, an oligonucleotide complementary to the
target sequences was utilized. This oligonucleotide hybridized to BJhTert
and NIH3T3 cells before incubation with anti-histone H3 antibody and antirabbit PLA-probe. Subsequently, two circularization oligonucleotides were
allowed to hybridize and ligate, and the resulting DNA-circle was amplified
by RCA and detected by fluorescence-labeled detection oligonucleotides.
Genomic DNA-templated in situ PLA
The next approach was to use single stranded genomic Alu repeats as
templates for ligation of the circular probes. Cells were first incubated with
primary antibody and secondary PLA-probe as described for hybridization
based in situ PLA. Subsequently, the two circularization oligonucleotides
were hybridized directly to Alu-repeats at the genomic DNA and the
secondary PLA-probe, followed by ligation using T4 DNA ligase. The
resulting circular DNA molecule was amplified by RCA as described for
hybridization based in situ PLA.
Padlock probe based in situ PLA
In the final approach, we used a padlock probe modified with two internal
uracil-containing hairpin structures. The padlock probe was first hybridized
and ligated to the genomic DNA, using Amp-ligase at 45oC before primary
antibody and secondary PLA-probe were added. Treatment with UracilDNA-glycosylase (UNG)/EndoIV was then performed to degrade the
hairpin-structures of the padlock probe, allowing it to hybridize to the PLAprobe. Upon hybridization of an additional splint oligonucleotide to the
PLA-probe, a circular DNA molecule could be created, amplified by RCA
and detected by fluorescence-labeled detection oligonucleotides (Figure 11).
38
Figure 11. Hairpin-containing padlock
probe.
The hairpin-containing padlock probe
hybridizes to the target sequence via the
target complementary sites of the probe
(green) and is ligated. Subsequently, the
hairpin structures (orange) can be resolved
by removing the uracil bases. Now the
parts of the probe complementary to the
antibody-bound PLA-probe (blue) are liberated and can hybridize to the PLA-probe
forming a circular DNA molecule.
Results
Hybridization based in situ PLA
To test if the selectivity of hybridization would be sufficient to discriminate
between human and mouse sequences, binding of the Alu-specific probe
alone was detected by a padlock probe hybridizing to the free 3’ end of the
PLA-probe. Even though the staining was more intense in human cells, there
was still substantial staining found in mouse cells, indicating unspecific
hybridization. The primary antibody was tested by IF and was found to give
a similar staining in both human and mouse cells.
When using probes for in situ PLA to detect protein-DNA interactions,
we observed ~160 interactions per human cell which was a 5-fold increase
compared to the mouse cells, where ~30 interactions per cell were detected.
Genomic DNA-templated in situ PLA
Since detection of target DNA by hybridization was not specific enough,
genomic DNA should be tested directly as PLA-probe. After incubation with
primary antibody and secondary PLA-probe, the circularization
oligonucleotides could hybridize to both the PLA-probe and genomic DNA,
and then be ligated. An additional tag oligonucleotide was incorporated into
the final circle to avoid unspecific ligation between two genomic sequences,
i.e. increasing the requirement for a perfect match at the PLA-probe. This tag
oligonucleotide served as a second detection site in the final RCP and thus
only double colored RCPs were regarded as true signals. With this second
design ~13 times more signals were detected in human compared to mouse
cells.
Padlock probe based in situ PLA
To further improve the selectivity of the DNA detection we switched ligase
to Amp-ligase that tolerates higher temperatures for ligation and thus is more
specific than T4 DNA ligase. Since the antibodies utilized for detection of
the protein part might not endure temperatures exceeding 37oC, the DNA
39
had to be targeted prior to the protein. Circularization oligonucleotides were
converted into a padlock probe to couple both sequences together and with
that increase the likelihood that both would bind to the same target sequence.
This necessitated however that the padlock probe contained sequences that
were complementary to the oligonucleotide carried by the PLA-probe.
Hence, the PLA-probe would be attracted to the site where the padlock probe
had bound without the requirement to recognize a primary antibody in its
proximity. To circumvent this problem, two hairpin structures were
introduced in the padlock probe, shielding the PLA-probe complementary
parts. Subsequent to ligation of the padlock probe to genomic DNA and
binding of the primary antibody and the secondary PLA-probe, the hairpin
structures in the padlock probe were opened. This was achieved by
degradation of the uracil residues incorporated in the padlock probe, hence
allowing the probe to hybridize to the PLA-probe. As the hairpin-containing
padlock probe could be amplified by RCA like a regular padlock probe, it
had to be ensured that the probe had been cleaved by UNG/EndoIV and
religated. We therefore introduced a tag oligonucleotide into the final
circular DNA (as described above) that served as a second detection site,
confirming that the padlock probe had been cleaved and religated. Only
double colored signals were regarded as true detected interactions whereof
~25 signals could be detected for interaction of Alu-repeats with histone H3
in human cells and almost no interactions were detected in mouse cells
(median = 0). Hence, padlock probe based in situ PLA resulted in a ~500
times higher signal in humans than in mice, showing the highest selectivity
of the methods tested herein.
Discussion
We have herein described the first method that enables a highly selective
detection of PDIs in situ, potent to discriminate between highly homologous
sequences. However, only a small fraction of all Alu repeats was detected,
giving a detection sensitivity of ~0.13%. Thus we are still far from being
able to detect transcription factors’ interaction with single copy genes. To
improve this method further and make it applicable for less abundant
interactions, accessibility of genomic DNA to the probe needs to be
improved without affecting the protein binding to DNA. To gain access to
genomic DNA, the nucleus had to be permeabilized and the degree of
permeabilization to gain access to the nucleus had to be balanced to the
amount of protein that was lost due to permeabilization. With padlock
probes, ~0.13% of all Alu-repeats could be detected, which suggests that a
great part of the genomic DNA is not accessible to restriction digestion and
exonuclease treatment or inaccessible for the probes.
40
Summary and future perspective
Analyses on a single cell level are essential to determine cell-to-cell
variations in a population and assays where this can be performed in
complex material, such as tissue sections, are needed for the methods to also
be applicable in a diagnostic setting. The papers presented in this thesis all
utilize in situ PLA in one way or another to investigate interacting
biomolecules and their modifications, as a measurement of protein activity
status. In paper I, we developed secondary PLA-probes to detect the
phosphorylation of individual PDGFR molecules. These probes allowed us
to apply primary antibodies, without the need for conjugation, as long as
these antibodies were derived from different species. Individual in situ PLA
signals could be quantified and with that provide a measurement of how
many receptors were phosphorylated upon different treatments. To increase
the throughput, we investigated in paper II if in situ PLA would be
compatible with flow cytometer readout for detection of homo- and
heterodimers of EGFR and HER2, as well as phosphorylation of EGFR. In
addition, we could also show that in situ PLA was superior to IF for
detection of phosphorylated EGFR due to its dual recognition. In paper III,
we combined in situ PLA detection of PDGFR phosphorylation with
padlock probe detection of DUSP6 mRNA, both being induced by PDGFBB stimulation. The ability to simultaneously measure activity status at
different nodes in a signaling cascade provides a tool to study kinetics at a
single cell level and can be utilized to dissect the molecular effect of drug
candidates. In paper IV, we developed a method for detection of PDI in situ
utilizing both in situ PLA and padlock probes, providing a highly selective
method and paving the way for future developments of assays to determine
protein interactions with single copy genes. The papers included in this
thesis are a contribution to the molecular toolbox offering means to deepen
our understanding of how cells are regulated through interactions between
various biomolecules and of what variation that occurs within a seemingly
homogenous cell population.
Of course, still too many questions remain unanswered, and are not
possible to answer with current available methods or have not yet been
asked. Therefore, new methods and improvements of existing methods are
always required. To retrieve as much information from each investigated cell
as possible, not only should the type of binders that can be used with in situ
PLA be expanded, but also the number of molecules that can be investigated
41
simultaneously. Therefore a multiplex version of in situ PLA has been
developed to determine the balance between alternative interaction partners
and to follow signaling pathways (Leuchowius et al., submitted).
Furthermore, the dynamic range of in situ PLA’s single molecule
discrimination is restricted since it has to be possible to distinguish
individual signals. To expand the dynamic range and dissolve signals that
have converged, a novel oligonucleotide design has been tested that can
increase the dynamic range by several orders of magnitude (Clausson et al.
in preparation). Also, other designs that provide the ability to decrease or
increase the distance requirement for the generation of in situ PLA signals
have been developed (Leuchowius et al., submitted). Utilization of screening
and scanning devices as well as flow cytometers makes in situ PLA
applicable even for high content analyses as demonstrated by Leuchowius et
al.105. Now it is possible to screen new drugs not only for their effect on the
expression of a protein but also for their impact on the function, and the
analyses can be done either on cell lines or in patient cells to support the
choice of therapy.
Methods to detect protein-DNA interactions for single copy genes
interacting with rare proteins like transcription factors are a future goal. To
achieve this goal with in situ PLA, the greatest current obstacle is that
genomic DNA has to be made accessible to probes, while also the proteins
have to stay in place. Therefore fixation and permeabilization need to be
improved. Furthermore, all enzyme and binder based reactions need to be
performed with high efficiency. Whether an interaction is detectable will
also depend on the place in the nucleus where this interaction occurs, and
how condensed the chromatin is at this place. Interactions between other
types of biomolecules such as protein-RNA also require method
development. As the efficiency of the DNA ligases for a DNA/RNA hybrid
is very poor, our method for PDI detection described herein would require
development of a ligase efficient in ligating a DNA molecule templated by
RNA. Alternatively, if an RNA molecule can ligate on an RNA template, an
RNA polymerase (or reverse transcriptase) would be capable to perform an
RCA.
However, despite all efforts to improve the sensitivity of the target
molecule detection – protein, nucleic acids and all combinations hereof – it
will be challenging to detect 100% of endogenously expressed, untagged
molecules present in a sample. What might be the greatest obstacles is the
accessibility of the target molecules and the fact that all types of binders and
enzymes have a limited affinity for their target but unfortunately also an
affinity for other, irrelevant molecules.
But where does that leave us? As mentioned in the beginning, in addition
to in vitro methods that provide important basic knowledge, methods that
visualize endogenous molecules in cells, tissues and whole animals will
contribute substantially to a deeper understanding of the processes of life. As
42
Richard P. Feynman said: “It is very easy to answer many of these
fundamental biological questions; you just look at the thing!”. Being able to
look at single molecules and how they interact in signaling cascades, in
individual cells and potentially in vivo, is desirable so that we one day can
understand how cells and whole organisms work. It is tempting to speculate
that methods could be developed to detect all proteins belonging to one
pathway and their interactions in multiplex, so that the signaling cascade
could literally be “looked at” in individual cells. If further protein detection
could be combined with multiplexed detection of RNA and DNA molecules
in situ, then gene expression profiles and mutations as well as changes in
signaling cascades could be studied in healthy and diseased cells. In that case
questions like: Where in the cell and tissue are the interactions taking place?
How often do certain interactions occur and how are they distributed in the
cell, the population, the tissue and the organism? – all leading up to the
ultimate question: Which molecular events cause or perpetuate a disease and
how can it be cured? – could be answered. The methods presented in this
thesis will hopefully be one step in the relay to answer these questions.
43
Acknowledgments
The work presented in this thesis was done at the Department of
Immunology, Genetics and Pathology at Uppsala University. First of all, I
would like to thank all organizations which have financially supported our
work, in particular: Beijerstiftelsen, Knut and Alice Wallenberg stiftelse, the
Swedish Foundation for Strategic Research, Vetenskapsrådet, Cancerfonden,
the National Biobank Platform, the EU-FP6 projects ENLIGHT and
MolTools, and the EU-FP7 project DiaTools. Without your support, this
research would not have been possible.
Then I would like to express my gratitude to my faculty opponent Prof.
Banafshe Larijani and the examination board Prof. Sophia Hober, Assoc.
Prof. Johan Lennartsson and Prof. Chandrasekhar Kanduri for taking the
time to read my thesis and for coming to Uppsala to discuss it with me.
I would like to thank all people, near and far, who have been at my side
during the last five years, in particular:
My supervisor Ola Söderberg, for introducing me to the wonderful world of
in situ PLA and the magic we can do with it. Thank you for all your support
during these years, for the open discussions we shared, for having a plan “B”
when I needed it, for showing me that one sometimes just has to give it a try
and for saving my life – from axe murderers on the plane and from drowning
in the pool when “swimming” butterfly.
My co-supervisor Ulf Landegren, for bringing me into the group, for always
having an open door and sharing your thoughts, no matter how busy you are,
for creating such an open atmosphere at the lab, which makes this group a
very special and inspiring place to do research in.
My collaborators on various projects: Thank you Carolina Wählby, for
opening my eyes to image analysis and for being such a wonderful person;
Milan Gavrilovic, for tireless explaining your algorithms and for counting
my yellow blobs; Amin Allalou, for writing Blobfinder, it made my life a lot
easier; Dirk “Pacho” Pacholsky and Jan Grawé, for keeping our
microscopes up and running and for being at call, when they would not
cooperate; Nongnit Laytragoon-Lewin, for sharing the cell lab with us;
44
Arne Östman, for your support on paper I and for always believing in
oxPTP-in situ PLA, I still believe that it should be possible; Janna
Paulsson, for the great work on paper I; Markus Dagnell, for your
engagement in the oxPTP project and all the rest of the Östman group, for
the great times I have had with you! Thank you Johan Kreuger and
Sébastien Le Jan for our work on embryoid bodies. Seb, I really enjoyed
our discussions during these looong hours at the microscope. Thanks to
Marianne Farnebo and Salah Mahmoudi for the collaboration on Wrap53.
A special thanks to Frank and Annette Böhmer for always asking the right
questions, for inspiring me with science, music and books and for bringing
in situ PLA to Jena.
Many thanks to the people from “my” molecular tools / molecular
medicine group (current and former) for making our lab such a fantastic
place to be at:
Ida, my second I2 half, thank you for being there for me during all these
years, for keeping the lab bench next to me, despite the RNases I worked
with, for introducing me to the gym and the sensation that having pain can
actually be fun, for teaching me Swedish and cross-country skiing, for your
company at various conferences, in short – for your friendship! Johan E, for
many evenings in your sofa discussing psychology and for your patience
when Ida and I were busy doing girly things.
Thanks to my in situ PLA family: Kalle, my brother in in situ PLA ;-), for
understanding my complaints about antibody-selectivity, for repeatedly
listening to my explanations about complicated protocols for oxPTP and
DNA-prot detection, for making Illustrator do what I want and for the great
fun we had outside the lab singing ”I can show you the world” and
“swimming” butterfly; Carl-Magnus, for staying in the swimming team,
when everyone else left and for enjoying non-black tea as much as I do;
Karin, for joining our in situ PLA group and girly orders from the US;
Agata, for enjoyable lab discussions and for trying to make people tell when
they “borrow” your stuff; Di, for our microscope sessions philosophizing
about Drosophila brains; Gucci, for introducing me to Chinese jazz music
and many books worth reading; Christina, for finally realizing that in situ
PLA is the right side of this lab, for testing cake recipes on us and for joining
in on countless gym classes; Tim, for asking the right questions and for
getting involved, for fixing my computer by just rolling over and for,
together with Lei and Louis, all the wonderful dinners at your place;
Katerina, for being such an amazing person and for bringing me down from
”Åreskutan” in one piece, you are a great skiing teacher!; Malin, for taking
me on as a student, for introducing me to in situ PLA and for valuable input
on defense forms; Georgia, for bringing back some Greek attitude into your
little in situ PLA group and for discussions about in situ PLA and life;
Andries, for chats about transfection and about how to make a bacterium do
45
what I want; Pier, for discussions about protein-DNA interactions and your
valuable help with the Bioanalyzer; Caroline and Mikaela, for just letting
me stop by and have a (in situ PLA) chat;
A special thanks to Lena, first of all for your help on paper IV, without you,
I would still stand in the lab and pipette! But thank you even more for your
friendship and the great times I got to spend with your family (Tack, Jesper
och Anton för mys- och filmkväller hemma hos er!); to Sara, for being my
”nucleus”-person, for introducing me to dancing Salsa and for all the
fabulous times we spend together; to Spyros, for sharing the experience of
the first Swedish winter, for the ”therapy” meetings at our office and for
being 100% committed as actor in all spex movies; to Mats for your
valuable input on padlocks, science in general and paper III in particular; to
Masood, for the introduction to Jujitsu, for explaining how to cook rice
properly and for your fabulous kind of humor; to Lore, for showing me how
I would like to be when I grow up: interested, engaged and caring.
I would like to thank all current and former lab members for sharing science
and parties at countless occasions:
Rachel, for your seemingly endless energy and gladness and for
refreshing discussions; Maria H, for great lunch and fika company and
excellent cake recipes; Anja, for evening discussions in German and for
offering to go climbing with me (when this is over, I will join you!); Elin
FS, for your help with counting sequences and for, together with Karin,
starting in the swimming team; Rongqin, for long-distance relationship
discussions at the microscope; Marco, for taking care of the microscopes
and for doing ”whatever you want” for Spyros movie; Johan V, for joining
the PhD-council and for visiting Jena; Junhong, for your company and great
dumplings at the Xmas party; Anne-Li, for always being positive and
engaged – in lab projects as well as spex movies, Ling, for patiently
listening to my explanations about cell culture; David, for trying to built a
“group memory”; Jenny, for your positive attitude and for knowing
everything about blobs; Mathias, for engaging in zinc fingers and for many
fun chats; Sigrun, for organizing the “come as you are not”-party together
with Jenny; Olle, for valuable input at group meetings, your contagious
laughter and for cocktail parties with “very interesting” cocktails; Henrik,
for entertaining chats in the lab and for the sandwiches you prepared for us
at 3:00 a.m. after a long night at BJ’s; Magnus and Anna, for your company
at grill parties; Yuki, for delicious sushi nights and for importing culinary
specialties from all over the world; Chatarina, for discussions about RGB
and CYMK colors, Jörg, for introducing me to DNA-protein interactions;
Ida E, for having such an irresistible laughter and Lotte, Artor, Joakim,
Camilla, Lei, Björn, Annika, Elin L, Anna E, Monica, Ellen, Jonas,
Johan S, Sigrun, Heidi, Mia, Carolina R, Edith, Reza and Anna K for
46
lively discussions and inspiring chats, for lunch and fika company and for
making this group such a special place.
A huge thank you to the people in the group who are taking on the
administration: Elin E, for taking care of all the orders and for being there
when I needed a break from writing; Carla, for your help with antibody
conjugation and both of you together with Lena for keeping the lab up and
running and us organized; Erik, for answering all my questions about
funding and for chairing this defense wearing a cerise colored shirt ;-); Ulla,
for chats early in the morning when I was scuffling to the microscope and
for patiently explaining in which order the registration forms should be
signed; Johan O, for rescuing my external hard drive even though you had
warned me that it was not secure; Anders, for having such a great kind of
humor and for choir talks over lunch.
I would like to thank my students Eric and Jane for their work in and
enthusiasm for the protein-DNA project and for believing in it as much as I
did.
The PhD student council, for all meetings and discussions about things we
were supposed to do and Rudbeck-life in general ;-).
People at Olink: Kicki, Mats, Göran, Daniel, Erik, Gabriella, Andrea,
Eva, Peter, Björn and the rest of the Olink team, for their help with
secondary in situ PLA, for countless “Duolink user” / Rudbeck people –
Olink meetings and for providing us with kits and special wishes throughout
the years.
I also would like to thanks some dear friends outside the lab: Zandra, for
giving me a place to live when I first came to Sweden, for our TV-nights,
delicious food and for always knowing how to do things; Anna, for singing
with me (and we will give that concert!), for explaining the Swedish pension
scheme and for our weekly therapy sessions in my sofa; Toni, for from time
to time reminding me what we dreamt of when we were little and for finally
visiting Uppsala; Maria S, for singing and playing the piano with me and for
walking BACK from Hammarskog as well, even though we did not get cake;
Antje, for our IKEA trip, for knowing who would win the Eurovision song
contest 2008 and for joining me swimming, even though you were in
Germany; ”die Mädels” – Annika, Berit, Bine, Susi and Janine, for
staying in touch and for the great reunion we had last year, let´s make it only
the first of many; Sina, for the countless cookies you baked for me, for long
discussions about science and life and for making me ski to Hammarskog.
47
”My” choir Collegium cantorum: Thank you for the music!!!, for listening
to my German without an ”R” and for the amazing trip to Iceland last
summer.
Ich möchte mich auch bei meiner Familie bedanken:
Bei meinen „Schwiegereltern” Konni und Frank sowie Christian, Oma
Käthe und Oma Renate dafür, dass ihr mich so lieb in die Familie
aufgenommen habt, für eure „Carepakete“ aus deutschen Landen und dass
ihr auf Basti aufgepasst habt.
Danke auch an meine Großeltern: Oma Rosi, dafür dass du mich dazu
gebracht hast, einmal darüber nachzudenken, was ich hier eigentlich mache,
und für deine Bedenken, ich könnte deutsch verlernen; Oma Inga und Opa
Jochen, für alle telefonischen Ferndiagnosen in den letzten Jahren und dass
ihr euch Sorgen über meinen Heimweg durch die „dunklen Wälder”
Uppsalas gemacht habt.
Opa, ich habe mir sehr gewünscht, dass du diese Arbeit noch lesen kannst,
hätte sie dir gern erklärt und mit dir darüber diskutiert, doch es hat nicht
sollen sein. Du fehlst mir!
Mama und Papa, danke, dass ihr immer für mich da seid, dass ihr euch
jederzeit ins Auto nach Uppsala gesetzt hättet, um mich zu retten, danke für
eure Liebe und dafür, dass ihr an mich glaubt! Vielen Dank auch an meine
Geschwister: Clausi dafür, dass du meinen Computer soo viele Male per
Ferndiagnose repariert hast, dass du zusammen mit Basti den Umzug nach
Schweden geschmissen hast und für die interessanten Gespräche, die wir auf
langen Autobahnfahrten und am Telefon gehabt haben. Danke auch an
Corinna dafür, dass du auf Clausi aufpasst und ans Telefon gehst, wenn er
mal wieder in der Firma verschollen ist. Lieschen, mein Wuschel, vielen
Dank für all die tollen Bilder (vor allem für das auf dem Umschlag!), damit
hast du mir immer eine große Freude gemacht.
Basti, mein Prinz, danke, dass du es trotz des großen Abstandes so lange mit
mir ausgehalten hast. Danke, dass du immer für mich da bist, auch wenn das
in letzter Zeit sicher nicht immer einfach war, für deinen Glauben an mich,
dafür, dass alles wieder gut wird, wenn ich mich bei dir anlehne und vor
allem für deine Liebe!
Was meinst du? Sollen wir es jetzt mal im selben Land, der selben Stadt, der
selben Wohnung versuchen? Ich liebe dich!
48
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Savage, D.C. Microbial ecology of the gastrointestinal tract. Annu
Rev Microbiol 31, 107-133 (1977).
Finishing the euchromatic sequence of the human genome. Nature
431, 931-945 (2004).
Misteli, T. Physiological importance of RNA and protein mobility in
the cell nucleus. Histochem Cell Biol 129, 5-11 (2008).
Pierige, F., Serafini, S., Rossi, L. & Magnani, M. Cell-based drug
delivery. Adv Drug Deliv Rev 60, 286-295 (2008).
Levsky, J.M. & Singer, R.H. Gene expression and the myth of the
average cell. Trends Cell Biol 13, 4-6 (2003).
Park, J., et al. Single-molecule analysis reveals the kinetics and
physiological relevance of MutL-ssDNA binding. PLoS One 5,
e15496 (2010).
Uyemura, T., Takagi, H., Yanagida, T. & Sako, Y. Single-molecule
analysis of epidermal growth factor signaling that leads to
ultrasensitive calcium response. Biophys J 88, 3720-3730 (2005).
Greenleaf, W.J., Woodside, M.T. & Block, S.M. High-resolution,
single-molecule measurements of biomolecular motion. Annu Rev
Biophys Biomol Struct 36, 171-190 (2007).
Monti, M., Orru, S., Pagnozzi, D. & Pucci, P. Interaction
proteomics. Biosci Rep 25, 45-56 (2005).
Fields, S. & Song, O. A novel genetic system to detect proteinprotein interactions. Nature 340, 245-246 (1989).
Aronheim, A., Zandi, E., Hennemann, H., Elledge, S.J. & Karin, M.
Isolation of an AP-1 repressor by a novel method for detecting
protein-protein interactions. Mol Cell Biol 17, 3094-3102 (1997).
MacBeath, G. & Schreiber, S.L. Printing proteins as microarrays for
high-throughput function determination. Science 289, 1760-1763
(2000).
Zhu, H., et al. Global analysis of protein activities using proteome
chips. Science 293, 2101-2105 (2001).
[Anon].
Identification
of
associated
proteins
by
coimmunoprecipitation. Nature Methods 2, 475-476 (2005).
Johnsson, N. & Varshavsky, A. Split ubiquitin as a sensor of protein
interactions in vivo. Proc Natl Acad Sci U S A 91, 10340-10344
(1994).
Fredriksson, S., et al. Protein detection using proximity-dependent
DNA ligation assays. Nat Biotechnol 20, 473-477 (2002).
49
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
50
Berggard, T., Linse, S. & James, P. Methods for the detection and
analysis of protein-protein interactions. Proteomics 7, 2833-2842
(2007).
Vogel, S.S., Thaler, C. & Koushik, S.V. Fanciful FRET. Sci STKE
2006, re2 (2006).
Sorkin, A., McClure, M., Huang, F. & Carter, R. Interaction of EGF
receptor and grb2 in living cells visualized by fluorescence
resonance energy transfer (FRET) microscopy. Curr Biol 10, 13951398 (2000).
Kenworthy, A.K. Imaging protein-protein interactions using
fluorescence resonance energy transfer microscopy. Methods 24,
289-296 (2001).
Sako, Y., Minoghchi, S. & Yanagida, T. Single-molecule imaging of
EGFR signalling on the surface of living cells. Nat Cell Biol 2, 168172 (2000).
Galperin, E., Verkhusha, V.V. & Sorkin, A. Three-chromophore
FRET microscopy to analyze multiprotein interactions in living
cells. Nat Methods 1, 209-217 (2004).
Xu, Y., Piston, D.W. & Johnson, C.H. A bioluminescence resonance
energy transfer (BRET) system: application to interacting circadian
clock proteins. Proc Natl Acad Sci U S A 96, 151-156 (1999).
De, A., Loening, A.M. & Gambhir, S.S. An improved
bioluminescence resonance energy transfer strategy for imaging
intracellular events in single cells and living subjects. Cancer Res
67, 7175-7183 (2007).
Xu, X., et al. Imaging protein interactions with bioluminescence
resonance energy transfer (BRET) in plant and mammalian cells and
tissues. Proc Natl Acad Sci U S A 104, 10264-10269 (2007).
Hu, C.D., Chinenov, Y. & Kerppola, T.K. Visualization of
interactions among bZIP and Rel family proteins in living cells
using bimolecular fluorescence complementation. Mol Cell 9, 789798 (2002).
Ciruela, F. Fluorescence-based methods in the study of proteinprotein interactions in living cells. Curr Opin Biotechnol 19, 338343 (2008).
Kerppola, T.K. Visualization of molecular interactions by
fluorescence complementation. Nat Rev Mol Cell Biol 7, 449-456
(2006).
Kerppola, T.K. Complementary methods for studies of protein
interactions in living cells. Nat Methods 3, 969-971 (2006).
Stolpe, T., et al. In planta analysis of protein-protein interactions
related to light signaling by bimolecular fluorescence
complementation. Protoplasma 226, 137-146 (2005).
Sung, M.K. & Huh, W.K. Bimolecular fluorescence
complementation analysis system for in vivo detection of proteinprotein interaction in Saccharomyces cerevisiae. Yeast 24, 767-775
(2007).
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
Hu, C.D. & Kerppola, T.K. Simultaneous visualization of multiple
protein interactions in living cells using multicolor fluorescence
complementation analysis. Nat Biotechnol 21, 539-545 (2003).
Shyu, Y.J., Suarez, C.D. & Hu, C.D. Visualization of AP-1 NFkappaB ternary complexes in living cells by using a BiFC-based
FRET. Proc Natl Acad Sci U S A 105, 151-156 (2008).
Gandia, J., et al. Detection of higher-order G protein-coupled
receptor oligomers by a combined BRET-BiFC technique. FEBS
Lett 582, 2979-2984 (2008).
Carriba, P., et al. Detection of heteromerization of more than two
proteins by sequential BRET-FRET. Nat Methods 5, 727-733
(2008).
Soderberg, O., et al. Direct observation of individual endogenous
protein complexes in situ by proximity ligation. Nat Methods 3, 9951000 (2006).
Gullberg, M., et al. Cytokine detection by antibody-based proximity
ligation. Proc Natl Acad Sci U S A 101, 8420-8424 (2004).
Lizardi, P.M., et al. Mutation detection and single-molecule
counting using isothermal rolling-circle amplification. Nat Genet 19,
225-232 (1998).
Baner, J., Nilsson, M., Mendel-Hartvig, M. & Landegren, U. Signal
amplification of padlock probes by rolling circle replication. Nucleic
Acids Res 26, 5073-5078 (1998).
Zieba, A., et al. Bright-field microscopy visualization of proteins
and protein complexes by in situ proximity ligation with peroxidase
detection. Clin Chem 56, 99-110 (2010).
Sigal, A., et al. Variability and memory of protein levels in human
cells. Nature 444, 643-646 (2006).
Lahav, G., et al. Dynamics of the p53-Mdm2 feedback loop in
individual cells. Nat Genet 36, 147-150 (2004).
Raj, A., Peskin, C.S., Tranchina, D., Vargas, D.Y. & Tyagi, S.
Stochastic mRNA synthesis in mammalian cells. PLoS Biol 4, e309
(2006).
Dedon, P.C., Soults, J.A., Allis, C.D. & Gorovsky, M.A. A
simplified formaldehyde fixation and immunoprecipitation
technique for studying protein-DNA interactions. Anal Biochem 197,
83-90 (1991).
Ren, B., et al. Genome-wide location and function of DNA binding
proteins. Science 290, 2306-2309 (2000).
Orlando, V. Mapping chromosomal proteins in vivo by
formaldehyde-crosslinked-chromatin immunoprecipitation. Trends
Biochem Sci 25, 99-104 (2000).
Kim, T.M. & Park, P.J. Advances in analysis of transcriptional
regulatory networks. Wiley Interdiscip Rev Syst Biol Med 3, 21-35
(2011).
O'Neill, L.P., VerMilyea, M.D. & Turner, B.M. Epigenetic
characterization of the early embryo with a chromatin
51
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
52
immunoprecipitation protocol applicable to small cell populations.
Nat Genet 38, 835-841 (2006).
Gustafsdottir, S.M., et al. In vitro analysis of DNA-protein
interactions by proximity ligation. Proc Natl Acad Sci U S A 104,
3067-3072 (2007).
Cremazy, F.G., et al. Imaging in situ protein-DNA interactions in
the cell nucleus using FRET-FLIM. Exp Cell Res 309, 390-396
(2005).
Zirbel, R.M., Mathieu, U.R., Kurz, A., Cremer, T. & Lichter, P.
Evidence for a nuclear compartment of transcription and splicing
located at chromosome domain boundaries. Chromosome Res 1, 93106 (1993).
Brown, K. Visualizing nuclear proteins together with transcribed
and inactive genes in structurally preserved cells. Methods 26, 10-18
(2002).
Brown, K.E., et al. Association of transcriptionally silent genes with
Ikaros complexes at centromeric heterochromatin. Cell 91, 845-854
(1997).
Zinner, R., Teller, K., Versteeg, R., Cremer, T. & Cremer, M.
Biochemistry meets nuclear architecture: multicolor immuno-FISH
for co-localization analysis of chromosome segments and
differentially expressed gene loci with various histone methylations.
Adv Enzyme Regul 47, 223-241 (2007).
Gall, J.G. & Pardue, M.L. Formation and detection of RNA-DNA
hybrid molecules in cytological preparations. Proc Natl Acad Sci U
S A 63, 378-383 (1969).
Levsky, J.M. & Singer, R.H. Fluorescence in situ hybridization:
past, present and future. J Cell Sci 116, 2833-2838 (2003).
Speicher, M.R. & Carter, N.P. The new cytogenetics: blurring the
boundaries with molecular biology. Nat Rev Genet 6, 782-792
(2005).
Schrock, E., et al. Multicolor spectral karyotyping of human
chromosomes. Science 273, 494-497 (1996).
Speicher, M.R., Gwyn Ballard, S. & Ward, D.C. Karyotyping
human chromosomes by combinatorial multi-fluor FISH. Nat Genet
12, 368-375 (1996).
Karhu, R., et al. Chromosome arm-specific multicolor FISH. Genes
Chromosomes Cancer 30, 105-109 (2001).
Femino, A.M., Fay, F.S., Fogarty, K. & Singer, R.H. Visualization
of single RNA transcripts in situ. Science 280, 585-590 (1998).
Raj, A., van den Bogaard, P., Rifkin, S.A., van Oudenaarden, A. &
Tyagi, S. Imaging individual mRNA molecules using multiple
singly labeled probes. Nat Methods 5, 877-879 (2008).
Silahtaroglu, A.N., et al. Detection of microRNAs in frozen tissue
sections by fluorescence in situ hybridization using locked nucleic
acid probes and tyramide signal amplification. Nat Protoc 2, 25202528 (2007).
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
Lu, J. & Tsourkas, A. Imaging individual microRNAs in single
mammalian cells in situ. Nucleic Acids Res 37, e100 (2009).
Volpi, E.V. & Bridger, J.M. FISH glossary: an overview of the
fluorescence in situ hybridization technique. Biotechniques 45, 385386, 388, 390 passim (2008).
Dirks, R.W. & Tanke, H.J. Advances in fluorescent tracking of
nucleic acids in living cells. Biotechniques 40, 489-496 (2006).
Itzkovitz, S. & van Oudenaarden, A. Validating transcripts with
probes and imaging technology. Nat Methods 8, S12-19 (2011).
Koch, J.E., Kolvraa, S., Petersen, K.B., Gregersen, N. & Bolund, L.
Oligonucleotide-priming methods for the chromosome-specific
labelling of alpha satellite DNA in situ. Chromosoma 98, 259-265
(1989).
Pellestor, F. Development and adaptation of the PRINS technology:
an overview. Methods Mol Biol 334, 211-220 (2006).
Kadandale, J.S., Tunca, Y. & Tharapel, A.T. Chromosomal
localization of single copy genes SRY and SOX3 by primed in situ
labeling (PRINS). Microb Comp Genomics 5, 71-74 (2000).
Troyer, D.L., Xie, H. & Hu, J. Direct in situ single-copy (DISC)PCR. Methods Mol Biol 71, 71-76 (1997).
Bagasra, O. Protocols for the in situ PCR-amplification and
detection of mRNA and DNA sequences. Nat Protoc 2, 2782-2795
(2007).
Player, A.N., Shen, L.P., Kenny, D., Antao, V.P. & Kolberg, J.A.
Single-copy gene detection using branched DNA (bDNA) in situ
hybridization. J Histochem Cytochem 49, 603-612 (2001).
Landegren, U., Kaiser, R., Sanders, J. & Hood, L. A ligase-mediated
gene detection technique. Science 241, 1077-1080 (1988).
Nilsson, M., et al. Padlock probes: circularizing oligonucleotides for
localized DNA detection. Science 265, 2085-2088 (1994).
Nilsson, M., et al. Padlock probes reveal single-nucleotide
differences, parent of origin and in situ distribution of centromeric
sequences in human chromosomes 13 and 21. Nat Genet 16, 252255 (1997).
Larsson, C., et al. In situ genotyping individual DNA molecules by
target-primed rolling-circle amplification of padlock probes. Nat
Methods 1, 227-232 (2004).
Shaposhnikov, S., Larsson, C., Henriksson, S., Collins, A. &
Nilsson, M. Detection of Alu sequences and mtDNA in comets
using padlock probes. Mutagenesis 21, 243-247 (2006).
Larsson, C., Grundberg, I., Soderberg, O. & Nilsson, M. In situ
detection and genotyping of individual mRNA molecules. Nat
Methods 7, 395-397 (2010).
Hart, C.E., et al. Two classes of PDGF receptor recognize different
isoforms of PDGF. Science 240, 1529-1531 (1988).
Bergsten, E., et al. PDGF-D is a specific, protease-activated ligand
for the PDGF beta-receptor. Nat Cell Biol 3, 512-516 (2001).
53
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
54
Ostman, A. PDGF receptors-mediators of autocrine tumor growth
and regulators of tumor vasculature and stroma. Cytokine Growth
Factor Rev 15, 275-286 (2004).
Heldin, C.H., Ostman, A. & Ronnstrand, L. Signal transduction via
platelet-derived growth factor receptors. Biochim Biophys Acta
1378, F79-113 (1998).
Baselga, J. Targeting tyrosine kinases in cancer: the second wave.
Science 312, 1175-1178 (2006).
Matsui, T., et al. Isolation of a novel receptor cDNA establishes the
existence of two PDGF receptor genes. Science 243, 800-804
(1989).
Matsui, T., et al. Independent expression of human alpha or beta
platelet-derived growth factor receptor cDNAs in a naive
hematopoietic cell leads to functional coupling with mitogenic and
chemotactic signaling pathways. Proc Natl Acad Sci U S A 86, 83148318 (1989).
Bermudez, O., Pages, G. & Gimond, C. The dual-specificity MAP
kinase phosphatases: critical roles in development and cancer. Am J
Physiol Cell Physiol 299, C189-202 (2010).
Jurek, A., Amagasaki, K., Gembarska, A., Heldin, C.H. &
Lennartsson, J. Negative and positive regulation of MAPK
phosphatase 3 controls platelet-derived growth factor-induced Erk
activation. J Biol Chem 284, 4626-4634 (2009).
Yarden, Y. The EGFR family and its ligands in human cancer.
signalling mechanisms and therapeutic opportunities. Eur J Cancer
37 Suppl 4, S3-8 (2001).
Schlessinger, J. Ligand-induced, receptor-mediated dimerization and
activation of EGF receptor. Cell 110, 669-672 (2002).
Moriki, T., Maruyama, H. & Maruyama, I.N. Activation of
preformed EGF receptor dimers by ligand-induced rotation of the
transmembrane domain. J Mol Biol 311, 1011-1026 (2001).
Bonasio, R., Tu, S. & Reinberg, D. Molecular signals of epigenetic
states. Science 330, 612-616 (2010).
Lander, E.S., et al. Initial sequencing and analysis of the human
genome. Nature 409, 860-921 (2001).
Deininger, P.L. & Batzer, M.A. Alu repeats and human disease. Mol
Genet Metab 67, 183-193 (1999).
Rudiger, N.S., Gregersen, N. & Kielland-Brandt, M.C. One short
well conserved region of Alu-sequences is involved in human gene
rearrangements and has homology with prokaryotic chi. Nucleic
Acids Res 23, 256-260 (1995).
Deininger, P.L., Jolly, D.J., Rubin, C.M., Friedmann, T. & Schmid,
C.W. Base sequence studies of 300 nucleotide renatured repeated
human DNA clones. J Mol Biol 151, 17-33 (1981).
Rodriguez, J., et al. Genome-wide tracking of unmethylated DNA
Alu repeats in normal and cancer cells. Nucleic Acids Res 36, 770784 (2008).
98.
99.
100.
101.
102.
103.
104.
105.
Sorkin, A., Eriksson, A., Heldin, C.H., Westermark, B. & ClaessonWelsh, L. Pool of ligand-bound platelet-derived growth factor betareceptors remain activated and tyrosine phosphorylated after
internalization. J Cell Physiol 156, 373-382 (1993).
Tidball, J.G. & Spencer, M.J. PDGF stimulation induces
phosphorylation of talin and cytoskeletal reorganization in skeletal
muscle. J Cell Biol 123, 627-635 (1993).
Rajkumar, V.S., et al. Platelet-derived growth factor-beta receptor
activation is essential for fibroblast and pericyte recruitment during
cutaneous wound healing. Am J Pathol 169, 2254-2265 (2006).
Nilsson, I., et al. VEGF receptor 2/-3 heterodimers detected in situ
by proximity ligation on angiogenic sprouts. EMBO J 29, 1377-1388
(2010).
Rasband, W.S. ImageJ. (1997-2005).
Carpenter, A.E., et al. CellProfiler: image analysis software for
identifying and quantifying cell phenotypes. Genome Biol 7, R100
(2006).
Krayev, A.S., et al. The nucleotide sequence of the ubiquitous
repetitive DNA sequence B1 complementary to the most abundant
class of mouse fold-back RNA. Nucleic Acids Res 8, 1201-1215
(1980).
Leuchowius, K.J., et al. High content screening for inhibitors of
protein interactions and post-translational modifications in primary
cells by proximity ligation. Mol Cell Proteomics (2009).
55